Benefit/cost ratio is 6 (from absenteeism) to 400 (from decreased disability). 7/
Verdict: the product is cost-effective, easy to scale, and does not crowd out evidence-based care. 8/
Labor-market outcomes (absenteeism, employment) increase by 0.1 SD 2/
Our findings suggest that early use led to skill adoption: thus, disengagement doesn't imply ineffectiveness; donors and int'nal orgs should consider stopping using engagement as a proxy for impacts. 5/
Effects persist despite high but rapidly decreasing app use (as common with digital products) 4/
App access _increases_ psychotherapy use; theory + evidence suggest that the app likely lowers cost/increases benefit of therapy. 6/
Three must read papers for PhD students. #scisky #PhD #science #research #academicsky
1. The importance of stupidity in scientific research
Open Access
journals.biologists.com/jcs/article/...
New WP w/ @raissafabregas @antoniaavazquez
An AI-powered mental health app improves mental health (depression, anxiety, stress, wellbeing) by 0.3 SD at 6 months, with other positive impacts: tiny.cc/rz71101 1/
No increase in severe depression and anxiety alleviates safety concerns 3/
